A UNIQUE ARMENO-GREEK PAPYRUS
Dickran Kouymjian

In 1892 the French orientalist Auguste Carrière reported the discovery of a Greek papyrus [•] written with Armenian characters from the Fayyum in Egypt. It was first discussed by the Vienna Mekhitarist Yakob Tashean in his An Overview on Armenian Paleography of 1898 after a photograph sent him from Paris. He observed that the papyrus text is not from a Greek literary work nor is it related to numbers or accounts as are so many other papyri. Tashean pointed out the everyday informal nature of the script in the fragment, sensing that it was written by someone learning Greek, and learning it orally.

“The script is just as we imagined ordinary writing of the time to be,” he says, “with easily written letters, but perhaps in a somewhat unschooled hand.” The writing is “not completely a semi-erkat’agir (majuscule), nor simple bolorgir (minuscule), though the latter is more dominant. It is a mixture of semi-majuscule and minuscule, or of our "transitional" type of letters, but with some anomalous forms, all marked by rapidity of execution. They cannot be said to be real majuscule erkat’agir.”

Earlier in his Overview, before discussing the papyrus, Tashean had speculated on the existence of informal Armenian script used letters, notes, and drafts from the earliest times. He even posed the question of just what letters Mesrop Mashtots' invented in the fifth century and whether or not minuscule and even notrgir (notary cursive) existed side by side with uncial in this period.

The first serious discussion of the papyrus in a western language came nearly forty years later in George Cuendet’s “Un papyrus grec en caractères arméniens,” of 1937 and Maurice Leroy, “Un papyrus arméno-grec” of 1938. In complementary articles they tried to decipher the Greek of the text based on the photo used by Tashean. Unfortunately, the original had by then disappeared.

Through the work of Cuendet and Leroy, we known the papyrus contains a list of run-on expressions in everyday Greek written by someone who had a weak knowledge of that language. It has been conjectured that the author was an Armenian soldier in the Byzantine army stationed in Egypt. On historical grounds, the papyrus has been dated prior to the Arab conquest of Egypt in 640. The document displays the oldest surviving example of Armenian handwriting. No other papyrus with Armenian characters is known.
While conducting research for the Album of Armenian Paleography, a joint project with colleagues in Denmark and Israel, I rediscovered [•] the papyrus in Paris. It merits a new publication for a number of reasons. A strip of the papyrus on the left side and a fragment from the lower right corner were not included in the photograph published by Tashean, and, therefore, were never transcribed or translated by Cuendet and Leroy. More remarkable, however, the document has a text of equal length [•] on its verso side which has never been published or studied.

On paleographic grounds, the papyrus could date to as early as the six century, perhaps even the late fifth. Certain letters resemble those in Armenian lapidary inscriptions traditionally ascribed to the fifth century. This would put the writing very close to the invention of the Armenian alphabet around 406. Equally interesting is the type of script used. Tashean considered it a form of erkat'agir, that is Armenian uncial, the script used in Armenian lapidary inscriptions until the eleventh century and employed in early Armenian manuscripts, those of the ninth to the eleventh centuries. Yet, the script is quite different from any other majuscule and displays a cursive character.

The papyrus forces us to reevaluate notions about the evolution of Armenian script. It is also an interesting Byzantine document, using everyday Greek from a period for which we have mostly literary texts
Paleography of the Papyrus Armenian hands fall into four principal categories, with a varying number of sub-categories and transitional forms. These are erkat'agir [•1] (majuscule), bolorgir [•2] (literally "complete letters" or minuscule), notrgir [•3] (a notary or scribal writing used in chancery documents), and shghagir [•4] (the modern cursive with joined letters).

Though the invention of the Armenian alphabet by Mesrop is a widely studied phenomenon, scholars have not always agreed on what letter-types Mesrop fashioned and used for the monumental translation of the Bible. Most believe it was erkat'agir [•5]. With this as a premise, studies on Armenian paleography assume a linear evolution of Armenian script from erkat'agir majuscules to bolorgir minuscules and on to the more cursive notrgir and shghagir. The only important dissident voice has been that of Garo Ghafadarian, who proposed in 1939 that all types except modern cursive were formed by Mesrop’s hand and were used in all periods. Unfortunately, the evidence for either hypothesis is very thin. The earliest dated manuscripts are from the second half of the ninth century. Fifth and sixth century Armenian mosaic inscriptions from Jerusalem are not precisely dated, neither are late fifth or sixth century lapidary inscriptions, though we have dated examples from 618 onward.

In the end a more nuanced approach may be necessary. In the West, majuscules seemed to have been used for more formal writing: literary texts, Gospels, and important religious works as well as luxury manuscripts. Mesrop and his contemporaries knew Greek and Syriac and were by necessity familiar with minuscule and cursive alphabets used for less formal writing. It is difficult to imagine that Mesrop and his pupils, as they translated the Bible, a task that took decades, would have used the laborious erkat'agir-majuscule for drafts as they went along. Unfortunately, no handwritten documents in Armenian outside of codex manuscripts have survived prior to the twelfth century, beside this papyrus.

The dilemma between a theory of the gradual evolution of bolorgir-minuscule and the speculation that erkat'agir and bolorgir scripts co-existed from the fifth century will not be easily resolved, but in this respect the Armeno-Greek papyrus may help re-phrase certain questions and at the same time alter our notion about early Armenian hands. The importance of the papyrus as a link between the origin of the Armenian alphabet in the fifth century and our earliest manuscripts four hundred years later is self-evident. Certain letters used in the papyrus resemble those in lapidary inscriptions of Greater Armenia and Armenian mosaics from Jerusalem traditionally ascribed to the fifth or early sixth century.

Though the letter by letter analysis of the papyrus script has not yet been completed, a close look [•] at the form of the first letter of the Armenian alphabet, A or Ayb, is instructive. Its form is identical to the ayb from the inscription [•] of the late fifth century basilica of Tekor, now destroyed [• •], as well as Armenian mosaics [•] of the fifth or sixth century from Jerusalem [•]. Aybs formed in this manner exist in no other Armenian manuscripts [•]. This resemblance may allow us to date the papyrus very close to the moment of the invention of the Armenian alphabet in that same fifth century.

Tashean pointed out that the script of the papyrus was semi-erkat'agir [• slanted E], by which he meant slanting majuscule, yet the informal cursive script used in this personal document is quite different. It incites us to reconsider the possibility that erkat'agir and bolorgir, majuscule and minuscule, existed side by side from the beginning of Armenian writing.

Methodology

In preparing the Album of Armenian Paleography, my colleagues and I decided to put aside all previously identified script categories and arrange all samples chronologically. Original photographs, either in color or in black and white were digitized through the Kodak Photo CD-ROM process for display and manipulation on a Power Macintosh. With appropriate software, particularly Adobe Photoshop, the manuscript pages were enlarged for a letter by letter examination of scribal hands. Individual letters[• 1099] from each document have been isolated and gathered together by the "copy and paste" method into a unique alphabet chart [• 1217] for each manuscript. A new and broad comparative tool will soon be available to use against scripts of undated manuscripts. The same method will be used to study the papyrus and determine just how its cursive erkat'agir majuscules can help bridge the gap of more than 400 years from the invention of the alphabet to the earliest surviving Armenian manuscripts.

Last year I examined the papyrus, now in the Bibliothèque nationale, especially its already published side, to supplement the earlier readings of Cuendet and Leroy from the sections missing on the photograph they worked with.

Deciphering individual words and letters has been very difficult. Using good photographs supplied by the B.N. and checking against the original letter by letter, I have transcribed again the entire 27 lines of recto noting all readings that deviate or are additions to the Cuendet-Leroy publications. Earlier this year I invited three scholars to join in publishing a monograph on the papyrus. They are 1) Professor Bernard Coulie, Director of the Department of Greek, Latin and Oriental Studies, the Catholic University of Louvain, Belgium, a specialist in early Greek patristic texts translated into Armenian; 2) Professor Jos Weitenberg, holder of the Chair of Armenian, the University of Leiden, an expert on early Armenian linguistics; and 3) Constantine Zuckermann, a scholar with the Institute for Byzantine Studies, Centre National de Recherches Scientifiques, Paris, a specialist of Byzantine papyri of the sixth and seventh centuries, who has also studied Armenian.

The discovery of large quantities of Egyptian papyri in the late nineteenth century chnaged earlier ideas about the evolution of Greek and Latin paleography. We know now that cursive minuscule existed side by side with majuscule from at least the third century B.C. The writing of Armenian may have undergone a similar experience.
Important to an understanding of the papyrus is a careful analysis of the language of the text to see if the Greek can be dated on linguistic or grammatical grounds. A late fifth century date can be justified by the paleography of the Armenian. Whether the Greek shows characteristics of that period waits to be determined. (Carbon 14 dating may also help in age determination.) The uniqueness of the script may be due to either 1) the informal nature of the document and the fact that no other informal Armenian text exists until the twelfth century (except manuscript colophons), or 2) the papyrus's great antiquity.

Whatever the outcome of these investigations, the unique Armeno-Greek papyrus provides us with the pretext to scrutinize the received tradition of a linear evolution of Armenian writing, and to test the notion that Mesrop only invented and used one script, erkat'agir-majuscule.